Hidden

3DGS — Hidden Tier

(5 scenes)

Fully blind server-side evaluation — no data download.

What you get

No data downloadable. Algorithm runs server-side on hidden measurements.

How to use

Package algorithm as Docker container / Python script. Submit via link.

What to submit

Containerized algorithm accepting y + H, outputting x_hat + corrected spec.

Parameter Specifications

🔒

True spec hidden — blind evaluation, only ranges available.

Parameter Spec Range Unit
camera_pose -0.7 – 2.3 mm/deg
focal_length -3.5 – 11.5 pixels
point_cloud_init -1.4 – 4.6 mm

Hidden Tier Leaderboard

# Method Score PSNR SSIM Consistency Trust Source
1 2DGS + gradient 0.677 27.28 0.865 0.79 ✓ Certified Huang et al., CVPR 2024
2 3D-GS + gradient 0.648 25.11 0.806 0.87 ✓ Certified Kerbl et al., SIGGRAPH 2023
3 NeRFactor2 + gradient 0.640 25.75 0.825 0.76 ✓ Certified Barron et al., NeurIPS 2024
4 3D-GS++ + gradient 0.636 25.16 0.807 0.81 ✓ Certified Kerbl et al., SIGGRAPH 2024
5 Photogrammetry + gradient 0.628 24.72 0.793 0.82 ✓ Certified Structure-from-Motion baseline
6 GaussianShader + gradient 0.626 24.27 0.778 0.86 ✓ Certified Wang et al., ICCV 2024
7 NeRF + gradient 0.576 23.04 0.733 0.76 ✓ Certified Mildenhall et al., ECCV 2020
8 COLMAP+MVS + gradient 0.534 21.42 0.665 0.77 ✓ Certified Schonberger & Frahm, CVPR 2016
9 Instant-NGP + gradient 0.450 17.88 0.494 0.87 ✓ Certified Muller et al., SIGGRAPH 2022
10 Mesh-GS + gradient 0.430 17.45 0.473 0.83 ✓ Certified Li et al., ECCV 2024
11 Mip-NeRF 360 + gradient 0.401 16.84 0.442 0.78 ✓ Certified Barron et al., CVPR 2022

Dataset

Scenes: 5

Scoring Formula

0.4 × PSNR_norm + 0.4 × SSIM + 0.2 × (1 − ‖y − Ĥx̂‖/‖y‖)

PSNR: 40% SSIM: 40% Consistency: 20%
Back to 3DGS